<<

xiv PROLOGUE and production team. A number of anonymous reviewers provided 1 Beginning Concepts invaluable suggestions for improvement of the original manuscript. Dianne Bradley, Chuck Cairns, Dana McDaniel, Lucia Pozzan, and Irina Sekerina have provided guidance in a number of areas. We have also benefited from being part of the psychoIinguistics community in and around the CUNY Graduate Center and Queens College. We are fortunate to have students and colleagues with expertise in some of the we have used in examples throughout the book. For their help with these, we thank Yukiko Koizumi, Ping Li, Shukhan Ng, Irina Sekerina, Amit Shaked, IgIika Stoyneshka, and F. Scott Walters. Our primary. goal is not to provide our readers with a great many facts about acquisition and use. As in all healthy empirical fields, data change with ongoing investigations. Instead, we hope to convey to our readers the amazing story of the unconscious processes that take place as use language. The Creativity of Language 2 Language as Distinct from . , Eva Fernandez and Communication 3 HelenCaims Some Characteristics of the Linguistic System 6 The Distinction between Descriptive and Prescriptive 7 The Universality of Human Language 10 lmplications for the Acquisition of Language 10 How Language Pairs Sound and 11 and 15 The Speech Signal and Linguistic 17 Origins of Contemporary 20 How This Book Is Organized 22 New Concepts 23 Study 23

Psycholinguistics is an interdisciplinary field of study in which the goals are to understand how people acquire language, how people use language to speak and understand one another, and how language is represented and processed in the brain. PsychoIinguistics is primarily a sub-discipline of and , but it is also related to developmental psychology, , , and speech . The purpose of this book is to introduce the reader to some of the central ideas, problems, and discoveries in contemporary 2 BEGINNING CONCEPTS BEGINNING CONCEPTS 3

psycholinguistics. In this chapter, we explore key concepts about of topics. Many mammals have sets of calls and cries, but they language that serve to distinguish it from related aspects of human can communicate only certain kinds of information, such as whether behavior and , and we identify the basic characteristics of danger is coming from the ground or the air, who is ready to mate, language as a system. We also provide a brief account of how where food is located, and so forth. The philosopher Bertrand Russell psycholinguistics emerged as a field of inquiry. once said, "No matter how eloquently a dog may bark, he cannot tell you his were poor but honest" (Gleason and Ratner 1993: 9). Language is so flexible that it not only allows people to say anything • The Creativity of Human Language they can think of; it also allows people to use language for a vast array of purposes. Language is used to communicate, to interact socially, to A good place to begin is by thinking about some of the unique features entertain, and to inform. All cultural institutions - schools, communi­ of human language. Language is a system that allows people immense ties, governments - depend upon language to function. Written and creativity. This is not the same creativity of people who write essays, audio-recorded language allows people to communicate and convey fiction, or poetry. Instead, this is the linguistic creativity that is com­ information - as well as interact and entertain - across vast spans of monplace to every person who knows a language. The creativity of space and time. It is probably the case that human dominance of the human language is different from the communication system of any planet has been possible because of the power of human language as a other animal in a number of respects. For one, speakers of a language medium for transmitting knowledge (Dennett 2009). can create and understand novel sentences for an entire lifetime. Consider the fact that almost every sentence that a person hears every day is a brand new event not previously experienced, but which can be • Language as Distinct from Speech, Thought, understood with little difficulty. Similarly, when speaking, people con­ and Communication stantly produce novel sentences with no conscious effort. This is true for every person who speaks or has ever spoken a language. We can Language is the primary communication system for the human extend this to every person who uses a signed language to species. In ordinary circumstances it is used to convey produce and comprehend novel sentences. through speech. It is a special system, however, that functions inde­ This remarkable ability to deal with novelty in language is possible pendently of speech, thought, and communication. Because one of the because every language consists of a of principles by which arbi­ main themes of this book is to identify the unique aspects of the human trary elements (the sounds of speech, the gestures of ) are lingUistic system, it might be helpful to distinguish between language combined into words, which in tum are combined into sentences. and the other systems with which it usually interacts: speech, thought, Everyone who knows a language knows a relatively small number of and communication. principles, a small number of sounds put together to create words, and Before we discuss those other systems, let us emphasize that here a large but finite vocabulary. This finite knowledge provides the person and throughout this book our discussion of human language includes who knows a language with infinite creativity. The set of possible sen­ the signed languages of the deaf, unless explicitly noted. Sign languages tences for a given language is infinite. Everyone who has ever lived and are just as structured as any spoken language and are just as capable of known a particular language has produced and heard a miniscule conveying an unlimited range of topics (as discussed in the previous subset of that infinite set. Knowledge of language confers upon every section). Sign languages also operate under principles distinct from person the creativity to produce an infinite number of novel sentences. thought and communication. What differs between signed and spoken When that knowledge is shared with others in a given language com­ languages is the transmission mode: gestural for the former and articu­ munity, speakers and hearers are able to produce and understand an latory-phonetic (speech) for the latter. indefinitely large number of novel sentences. Speech ought not to be confused with language, though speech is A second important kind of creativity humans possess is that we can indeed the most frequent mode for transmitting linguistic information. use language to communicate anything we can think of. No other Other modes for transmission include the gestures used in sign lan­ animal communication system affords its users such an unlimited range guage and the graphic representations used in . Later in this 4 BEGINNING CONCEPTS BEGINNING CONCEPTS 5

chapter (and later in the book), we will address the differences between impaired their language ability. Moreover, many animals can think but the signal (speech, signs, written ) and the abstract information cannot communicate using language. In the language pathologies, we carried by that signal, and we will demonstrate that producing or per­ observe pronounced mismatches between level of intellectual devel­ ceiving a speech signal is possible and efficient because of knowledge opment and linguistic ability. Specific language impairment (SU) is of language. For now, consider the "linguistic" abilities of parrots and not a rare disorder in children without any neurological or motor computers. Both can produce speech that might sound very human­ pathology. In children with SU, lags far behind like (promising new technologies are also able to create gestural that of their peers. While there are numerous cognitive deficits associ­ sequences, using computer-animated figures, in sign language). But ated with children with SLI, their non-verbal is within animal or computer-generated speech (or signing) differs from true normal range and their cognitive deficits are not sufficient to account hu~an in one crucial respect: it is not based on for their language disorder (Leonard 1998). The flip side of SLI is knowledge of language as a finite system that yields an infinite set of Williams Syndrome, a genetically based disorder causing severe retar­ possible sentences. Notice in particular that parrot and computer speech dation. Children with Williams Syndrome are deficient in many other will fail to be creative in the senses described above. aspects of cognition. While some aspects of their language are impaired Another mode for transmitting linguistic information is writing, but Oacobson and Cairns 2009), these children have surprisingly good lan­ writing is markedly different from both speaking and signing. Writing guage skills, in both vocabulary and in the ability to form grammatical systems are invented by people who already use language, so the cen­ sentences (Lenhoff et al. 1997). PatholOgies such as SLI and Williams tral difference is that writing is a cultural artifact, while speaking and Syndrome, that demonstrate a dissociation of language and general signing are biological; we will examine this point in more detail in intelligence, are of interest because they demonstrate the independ­ Chapter 3. Writing is always dependent on spoken language, though ence of language and thought. the connection differs from language to language. In some languages, The thoughts that people have are distinct from the language (or lan­ like English, the written symbols - also called graphemes - are linked guages) in which they encode them. Bilinguals can use either of their to the language's sound system (consonants, vowels); in other lan­ languages to transmit the thoughts they want to convey. It may be that guages, like Chinese, the symbols represent words. Writing has had a one of the languages of a given bilingual will have a richer vocabulary very different historical trajectory than speech: humans have been for conveying certain thoughts, as in the person who prefers to speak using spoken language to communicate for tens of thousands of years, about art in English and about soccer in Portuguese. Perhaps it is more while writing is a relatively new development, with the earliest exam­ convenient to convey information in one of the two languages; for ples dating back to only about 5,000 years ago. Children learn to speak example, memorizing word lists in one language will facilitate recall in spontaneously and without explicit instruction, yet require hours and that same language (Cabeza and Lennartson 2005). But neither of these hours of teaching and practice when they are to read and write. phenomena alters the basic point: when required to, bilinguals are able While all human communities have some form of spoken (or gestural) to convey any thought in either of their languages, or in both. This language, in the majority of the world's languages a has observation can be extended to all human languages, of which there are not been invented. It is important to remember that languages without dose to 7,000 (Ladefoged, Ladefoged, and Everett 1997; Gordon 2005): a writing system are no less complex than their counterparts with any thought can be conveyed in any human language. A corollary of standardized writing systems. The complexity and sophistication of all this is that any sentence in any human language can be translated into human languages is independent of whether speakers have developed any other, even by ordinary bilinguals, as opposed to experienced a way to write the languages down. translators or trained interpreters. It may take more than one sentence It is tempting to confuse thought and language, because we verbal­ to do the job, and the translation may not be as elegant as the original, ize our thoughts using language. The distinction between language but all languages possess an ability to formulate equivalent meanings and thought (or general intelligence) becomes dear when one consid­ with precision. Thus, one can think of general intelligence as the system ers the many kinds of individuals who can think but cannot communi­ responsible for generating the "language of thought" (Fodor 1975), and cate through language. Among these kinds of individuals are this in tum is translated into speech by our linguistic system, which we and people who suffer from neurological pathologies that have describe in the following section and, in more detail, in Chapter 2. 6 BEGINNING CONCEPTS BEGINNING CONCEPTS 7 Language is the primary communication system for human beings, but it is not the only way to communicate, so language can be distinguished from communication in general. Many forms of communication are not linguistic; these include non-verbal, mathematical, and aesthetic com­ munication through music or the visual arts. Frequently, language is not used to communicate or transfer information; language can be used wMeaning Signal aesthetically (consider poetry or song lyrics) or as a means to negotiate social interactions (consider how Yo, whassup/ might be the preferred F1sure 1.1 Language is a system that connects signals (the sound wave on the greeting in some contexts but quite inappropriate in others). One of the right, symbolizing speech) and meanings (the light bulb on the left, symboliz­ wonderful things about language is that it can be studied in many dif­ Ing an idea). In the figure, the signal is acoustic, a speech sound. The signal ferent ways. Its social, cultural, and aesthetic characteristics can be ana­ ('ould take on other forms (it could be written, it could be gestural). lyzed independently of one another. In psycho linguistics, however, researchers are primarily concerned with the underlying structure of language as a biologically based characteristic of humans, derived from grammar and . Knowledge of such a system will give a speaker the human neurolOgical organization and function; we come back to the ability to organize ideas into words and sentences, and sentences this topic in greater detail in Chapter 3. Human language is unique to Into sequences of sounds. This special kind of knowledge is called tacit human beings and its general Structure is universal to our species. All (or implicit) knowledge, to distinguish it from explicit knowledge, such and only humans have human language. These facts have profound AS your knowledge of a friend's telephone number. Tacit knowledge is implications for the way language is acquired by infants (see Chapter 4) represented in the brain and is put to use, in this case, in the production and for the way that language is produced (Chapter 5) and perceived And comprehension of sentences, but is not consciously available to the (Chapters 6, 7, and 8). Individual who possesses it.

• Some Characteristics of the LingUistic System • The Distinction between Descriptive and Prescriptive Grammar Language is a formal system for pairing signals with meanings (see Figure 1.1). This pairing can go either way. When people produce a The term grammar means something different to linguists than what it sentence, they use language to encode the meaning that they wish to means to language teachers. People who teach language are interested convey into a sequence of speech sounds. When people understand a In teaching a standardized use of language, the form of a language that spoken sentence, language allows them to reverse the process and is accepted in academic and business circles. We can refer to this type of decode a speaker's speech to recover the intended meaning. Obviously, language as conforming to prescriptive grammar. Knowing how to these activities depend upon the speaker and hearer sharing a common ndapt to the standard (prescribed) way of speaking or writing is very language: both must have the same linguistic system for pairing sound useful for people conducting a job or producing a formal and meaning. piece of writing. People who study language, in contrast, are interested The linguistic system that enables sound and meaning to be paired in what is called descriptive grammar, that is, the language system that contains a complex and highly organized set of principles and rules. underlies ordinary use. This is not an easy concept to grasp, so some These rules are ultimately the source for the infinite creativity of examples are in order. Many people who speak English - especially language because they describe (or generate) anyone of an infinite set young people or people talking in informal contexts - will say sentences of sentences. The set of rules that creates sentences in a language is a like the following: language's grammar, and the words of a language are its lexicon. Notice that this way of defining language is very specific about what it (1 ) Me and Mary went to the movies. means to know a language. Knowing a language involves knowing its (2) Mary and me went to the movies. 8 BEGINNING CONCEPTS BEGINNING CONCEPTS 9

These sentences are generated by a person's internalized grammar of groups not only pronounce words differently, but also have profound English, which licenses those constructions, but which would not gen­ and highly systematic lexical and syntactic differences from the erate an ungrammatical sentence like the following: transnational standard version of English, or from regional stand­ ards, like Standard American English or Standard British English. (3) *Me went to the movies. For instance, people from the south of the United States use the word l1!lrse, whereas people from the north use the word pocketbook to refer (The asterisk, *, indicates that the sentence is badly formed.) The use of to the same thing. A feature ofSouthern American Vernacular English me in subject position is possible in English only with a compound sub­ is "modal stacking/' such that it is perfectly grammatical to say the ject (me and Mary or Mary and me), not with a singular one. A person sentence in (4), in which the two modal might and should are who can say (1) and (2) but not (3) has a particular kind of grammar stacked. that a linguist would want to be able to describe. English teachers are not interested in describing the properties of (4) We might should pay our bills tonight. people's underlying ; they want instead to make sure that their students know that certain ways of saying things are not consid­ Different dialects - their distinguishing properties, their origins, and ered "correct English." The prescriptive rules of English grammar their development over time - are of great interest to linguists. require that I be used in subject position, whether it is singular (I went So-called "standard" English, spoken by people like network news­ to the movies) or compound (Mary and I went to the movies). (English casters who have been trained to use it, is considered to be the ideal teachers would further object to (1) because it is considered impolite form of the language, but it is actually spoken by very few people. The to place oneself before others.) Similarly, students are told that they fact is that most people speak some sort of non-standard of should say It is I and This is she rather than It's me or This is her. English, some coming closer than others to the idealized standard However, most people - including the occasional English teacher, in form. Linguists do not take a position on whether there should be a casual speech - say It's me and This is her. The grammar that people standard version of a language or on what form the prescriptive rules develop during is the (colloquial) grammar of of the grammar should take. Yet language with prescriptive grammar other members of their language community. In fact, when people are guiding usage in formal contexts is a fact of life in modem society. acquiring the bulk of their linguistic ability in their first language (or Since business and professional communities ascribe to the ideat most languages) - a process that lasts from birth until a is around 5 or people would be well advised to become consciously aware of the dif­ 6 years of age - they have not even heard of linguistic correctness. ferences between the colloquial version of English acquired naturally There can be many differences between the sentences generated by by children (the language that linguists are interested in describing) that colloquial grammar and those sentences dictated by prescriptive and the standardized form of the language that will get someone a grammar. For example, many people will answer the telephone with good job or an A+ on an essay exam. It is a mistake, however, to believe It's me or This is her, rather than It is I or This is she. It is interesting to that there is anything inherently better about the set of sentences note that learning the prescribed rules of usage for a particular lan­ acceptable based on the prescriptive grammar of a language compared guage is often a tedious and difficult process, and one that requires a to those sentences generated by the grammar acquired naturally and great deal of conscious attention as well as explicit instruction, in con­ unconsciously. Unfortunately, non-standard varieties of English are trast to the ease with which children acquire (implicitly and without generally stigmatized, even by the very people who speak those varie­ instruction) the rules for the language or languages they acquire early ties (Preston 1998), and are often mistakenly seen as reflecting lack of on in life. intelligence or . Yet all human languages have variations The issue of correctness also arises when one considers dialectal that extend across their speakers, so if one considers a naturally occur­ variation. English, like most languages, takes on many different ring linguistic characteristic to be good, any deviations from the lin­ forms; the language varies geographically, by class, and by ethnicity. guistic norm are wonderful or at the very least, normal. The point is People from different English-speaking countries, from different that linguists are interested in describing people's grammars and dia­ areas within these countries, and from different racial and ethnic lects, and psycholinguists are interested in understanding how those 10 BEGINNING CONCEPTS BEGINNING CONCEPTS 11

grammars are put to use in the production and comprehension of sen­ I!I to skills that are learned later in life, such as riding a bicycle or writ­ tences. Psycholinguists are not concerned with correctness or stand­ Ing. If a person does not know how to ride a bicycle, one does not ard forms. n~sume there is anything wrong with this person, only that the person has not been taught how to ride a bicycle. If a person is unable to talk ~r a child is unable to acquire language, then one assumes a basic • The Universality of Human Language pathology and seeks professional advice. The rapid, effortless, and natural acquisition of language by children is likely a result of the fact Linguists tend to refer to human language as a single entity, despite the that language is a faculty of the . As the brain develops, it fact that there are many different versions spoken by the thousands of organizes the language the child is exposed to in ways that are common different language communities around the world. The fact is that all to all humans. human languages are cut from the same mold: they are highly similar This picture is complicated somewhat by second language acquisi­ in their organization and in the abilities they confer on the people who tion after , because learning a language as a teenager or know them. All human languages have a grammar and a lexicon, which as an is perceived as being very difficult, especially compared to together allow the creation of an infinite set of sentences to convey any the ease with which we learned our first language. Indeed, learning a possible thought. The fact that all humans have languages of similar second language is a great deal of work, particularly when the learner organization and function strongly suggests that language is part of the lives in an environment in which the language is not spoken regularly. human biological endowment, as the communication systems of ani­ Certain aspects of a second language are quite difficult to master, pro­ mals are specific to their species. The universality of human language nunciation in particular. And when learning a second language, one's has profound consequences for the way psycholinguists analyze the first language sometimes seems to get in the way. Yet (adult) second human use of language. language learners go through similar developmental stages as do At the same time, linguists are interested in understanding what is (child) first language learners. Furthermore, many people acquire high specific and what is universal, not only about knowledge of language levels of competence in a second language without having been taught but also about the mechanisms that put that knowledge of language to explicitly. Underlying these abilities, therefore, is a system for acquir­ use. The majority of the world's popUlation is bilingual or multilingual, ing human language that is engaged fully during first language acqui­ and most of the world's children grow up in environments that expose !'Iition and again at least partially with exposure to a second language, them to multiple languages (Romaine 1995). These facts indicate that at any time within the lifespan of an individual. To account for the the mechanisms for representing and processing language can handle perceived differences between first and second language acquisition, efficiently more than one linguistic code. research has pointed to variable amounts of exposure - usually vastly more extensive for first language learners - as well as to factors that include the leamer's psycho-social proximity to the target language • lmplications for the Acquisition of Language culture. Also, some recent proposals link age effects in second language acquisition to the decline in memory abilities observed with aging An important area of psycholinguistics is language acquisition. Just as (Birdsong 2005). every human culture has at least one language, children in every cul­ ture acquire the grammar and lexicon of the language or languages in their environment and develop the ability to employ that linguistic • How Language Pairs Sound and Meaning knowledge in the production and comprehension of speech. Children do this without effort and without being taught. Just as there are pro­ In any human language, the principles and rules of the grammar organ­ found similarities among human languages, there are profound simi­ ize words from the lexicon into sentences used to convey meaning. larities in the way children everywhere acquire their native language Three kinds of rule systems make up a grammar. Phonological rules or languages. Language acquisition is more similar to the acquisition describe the sound patterns of the language; they are used to create of other skills that develop in early childhood, such as , than it individual words and are responsible for the rhythm and intonation 12 BEGINNING CONCEPTS BEGINN1NG CONCEPTS 13

of speech. Morphological rules and syntactic rules are involved in (5) The senators objected to the plans proposed by the generals. creating the structural organization of words and sentences, that is, the relationships between words and phrases in sentences. (Chapter 2 (6) The senators proposed the plans objected to by the generals. describes the basic operations of these various rule systems, as well as the organization of the lexicon.) It is a fundamental concept in psy~ The meaning of the sentence in (5) is quite different from that of (6), cholinguistics that the meaning of a sentence is a function of the even though the only difference is the position of the words objected to meaning of individual words and how those words are organized and proposed. Although both sentences contain exactly the same words, structurally. People are consciously aware of many elements of lan~ the words are structurally related to each other differently; it is those guage - like consonants or vowels, syllables, and words - but they differences in structure that account for the difference in meaning. The tend not to be aware of sentence structure. When one reads in the pop­ same ten words could be combined in such a way that they would have ular press that some subculture, like teenagers or video gamers, has a no structure and no meaning: different "language," it usually turns out that this "language" differs from English only in that it has some special vocabulary items or some (7) The to plans senators objected proposed the by generals the. specialized pronunciation features. People are probably not as aware of sentence structure as they are of sounds and words, because An unstructured collection of words does not convey meaning, and the sentence structure is abstract in a way that sounds and words are not. same collection of words can mean different thingS depending upon The acoustic signal of a recorded sentence has properties that reflect their organization. A person who knew only a lexicon, without a prin­ the consonants and vowels it carries (more on this phenomenon in cipled system to combine the words into sentences, could get some Chapter 5). Also, though they are not usually pronounced in isolation, ideas across, but would lack a system of sufficient precision to convey words are generally written with spaces around them in most of the more than just some simple thoughts. world's writing systems. In contrast to sounds and words, syntactic Another way to get a sense of how meaning depends upon sentence structure is not represented in the spoken or written signal. At the structure is to see how the same string of words in the same linear same time, sentence structure is a central aspect of every sentence. order can convey two different meanings, depending upon the abstract Though it has no physical reality, sentence structure has psychological structure assigned to them. Consider the structurally ambiguous reality: it must be represented by the speaker and recovered by the sentence in (8): hearer in order for the meaning of a sentence to be conveyed. In other words, the meaning of a sentence depends on the structural organiza­ (8) The man saw the with the binoculars. tion of the sentence's words. When a person sets out to learn a new language, something usually The sentence can mean either that the man saw the boy by means of the done in school, the task is frequently conceptualized as memorizing binoculars or that the man saw a boy who had the binoculars. Thus, new vocabulary. Language learners quickly realize, though, that struc­ with the binoculars is associated either with the saw or with the ture is just as important a feature of a new language as is its vocabulary. noun boy. Indeed, bilinguals usually have a better sense of language structure Figure 1.2 illustrates the structural differences associated with each than monolinguals, because they are accustomed to noticing that ambi­ of the two meanings of (8), using tree diagrams to spell out the struc­ guities in one language are not parallel in the other, for example, and tural (hierarchical) relationships between the words for the two mean­ that word-by-word translations usually do not work. All of this makes ings of the sentence. In the top tree in Figure 1.2, with the binoculars is a bilinguals more consciously aware of sentence structure than are mono­ prepositional phrase (PP) completely separate from the linguals. (NP) that contains the noun boy. In contrast, in the bottom tree, the PP We can appreciate the importance of sentence structure by looking at with the binoculars is grouped inside the NP that contains boy. The struc­ examples within a single language. For instance, in English, the same tures illustrated in Figure 1.2 reflect the difference in meaning that dis­ set of words can convey different meanings if they are arranged in dif­ tinguishes the two interpretations of the sentence, namely, with the ferent ways. Consider the following: binoculars tells us the instrument used by the man to see the boy (top tree), 14 BEGINNING CONCEPTS BEGINNING CONCEPTS 15

(a) S ~ • linguistic Competence and linguistic NP VP Perfonnance

A grammar and a lexicon are those components of language that {r,~ allow sounds and meanings to be paired. When people know a lan­ The man saw Oat N P NP guage, they know its grammar and its lexicon. This knowledge is I I I ~ the boy with Det N called linguistic competence. Linguistic competence is a technical I I term, different from the usual meaning of the word competence. Being the binoculars competent at something usually means that a person has adequate (b) ~ abilities to perform an action with skill, but that is not what is meant by linguistic competence. Linguistic competence has no evaluative NP/' "vp~ connotation; it simply refers to the knowledge of language that is in a ~ ~ ~p person's brain (or mind), knowledge that provides a system for pair­ ing sound and meaning. Linguistic performance, in contrast, is the De'i r.~ use of such knowledge inthe actual processing ofsentences, by which I uw" ~ we mean their production and comprehension. Typically, linguists The man / '" P NP are concerned with deSCribing linguistic competence and psycholin­ OJ i I ~ guists are concerned with describing linguistic performance. Beyond the boy with Del N basic sentence processing, psycholinguists are also concerned with I I the actual use of language. After a sentence is processed, it is stored in the binoculars memory and combined with other sentences to form conversations and narratives. The description of how language is actually used is Figure 1.2 Abstract structures associated with the two meanings of the struc­ called , a topic we address in Chapter 8. It is important to turally ambiguous sentence The man saw the boy with the binoculars. Focus on the distinguish between the grammatical and pragmatic aspects of a par­ different location for the prepositional phrase (the shaded node labeled PP), ticular linguistic event. For example, let us return to the structurally with the binoculars, in each of the two structures. ambiguous sentence in (8). The sentence can have two distinct mean­ ings, each of which is described by a different structural representa­ or conveys information about which boy was seen, namely the one with tion, like those shown in Figure 1.2. These two structures are made binoculars (bottom tree). The crucial difference is that the node labeled available by the grammar and conform to a number of syntactic rules. PP (which dominates the prepositional phrase, with the binoculars) If this sentence is actually used by a speaker and understood by a attaches directly to the VP node in the top tree, but to the NP node in hearer, only one of the two meanings will be the one intended by the the bottom tree. speaker and only one of the two meanings (hopefully the same one!) The structures in Figure 1.2, like the ones that will appear elsewhere will be recovered by the hearer. Which meaning is intended or recov­ in this book, are not constructed with the type of detail a linguist would ered will be a purely pragmatic issue, determined by the situation, use. When linguists draw representations of the structures of a sen­ the participants in the conversation, the function of the communica­ tence, such theoretical objects take on a level of detail- like a drawing tive exchange, and so on. The grammar is completely indifferent to of a molecular structure by a biochemist - that goes well beyond our the speaker's intent or to the hearer's recovery of the message. The needs in this book. We will use simplified graphic representations, grammar simply provides structures that are available for the illustrating only the particular aspects ofsentence structure that need to encoding of meaning in sentences. The actual use of those sentences be focused on. The structural elements in Figure 1.2 will be described in in conversation is a function of encoding and decoding processes more detail in Chapter 2. and pragmatics. 16 BEGINNING CONCEPTS BEGINNING CONCEPTS 17

Like the complex processes underlying most of the activities of living­ walking, breathing, sleeping - the activities involved in the production and perception of sentences are completely unconscious. It is not pos­ sible to introspect and experience a piece of theprocess, like the retrieval w of words from the lexicon or the use of one's grammar to create a struc­ tural representation of a sentence. As we will see - particularly in Chapters 5, 6, and 7 - psycholinguists have developed experimental procedures that have led to an understanding of a great deal about Figure 1.3 Steps involved in encoding by the speaker (left to right) and decod­ these unconscious processes, which are quite remarkable in their speed ing by the hearer (right to left). and complexity. In the encoding process, an abstract object - an idea - is translated into a phYSical object - a speech signal. When we say that an idea is There are several actual processes that must take place when people abstract, we mean that it does not have an observable physical reality. use language to exchange ideas, processes for the production and per­ Certainly, an idea must have a physical representation deep in the neu­ ception of sentences. Figure 1.3 illustrates these operations, byexpand­ rological connections of the brain, but it has no such physical represen­ ing on the gray box of Figure 1.1. The speaker begins (top right of the tation for the hearer nor is that neurological representation measurable figure) with an idea or a thought she wants to convey to the hearer. In with ordinary instruments. Speech, on the other hand, is concrete; it is order to do this, she first must translate her thought into a semantic partof observable physical reality. Not only does it have an effect on the representation (a representation of meaning) for a sentence in her lan­ auditory system; it can also be recorded and its physical properties guage. Then she must select the words from her lexicon and use her measured. When the hearer decodes the physical Signal, he recovers grammar to construct the syntactic representation (representation of the same abstract object - the idea - that was encoded by the speaker. sentence structure) that will convey the meaning she has selected. The Let us take this a step further by pointing out that, since the idea and words must then be represented as sounds, that is, as a phonological the physical signal are not part of the linguistic system, neither is representation, since they are eventually going to be pronounced. directly reflected in the (also abstract) representations built by the lin­ Finally, the phonological representation is sent to the motor areas of guistic system during the encoding or decoding processes. (We come the speaker's brain and instructions are sent to the articulatory organs back to the nature of these abstract representations - the gray box of that are used to produce speech. The speech signal is the result of a Figure 1.1 in Chapter 2 and in Chapters 5,6, and 7.) The linguistic precisely timed and exquisitely organized interaction of hundreds of system is the system that bridges the idea and the speech, allOWing muscles, including those of the jaw, lips, tongue, vocal folds, and respi­ them to be related. The linguistic system represents sounds and words, ratory system. Speech sounds reach the auditory system of the hearer, and creates the structures that organize those sounds and words into and he begins the process of reconstruction that is necessary to decode sentences. the speaker's message. First, he must reconstruct the phonological rep­ resentation in order to recover the speaker's words and their mean­ ings. Then, using the grammatical and lexical knowledge that he shares • The Speech Signal and Linguistic Perception with the speaker, he must reconstruct the words' structural organiza­ tion. He then has sufficient information to recover the basic meaning The fact that the signal is the only physical link between speaker and for the speaker's sentence that will ultimately lead to her idea or hearer is a critical psycholinguistic point. The speech signal must con­ thought. (We have arbitrarily chosen to refer to the speaker as a woman, tain enough information for the hearer to reconstruct the abstract and to the hearer as a man. This is a convention we will follow through­ structures that eventually convey the abstract ideas, and that recon­ out the book.) struction is essential to the decoding process. To fully appreciate the Exchanging ideas using speech is so commonplace that people never complexity of this task, it is necessary to understand the relationship think about the complex cognitive processes that underlie that experience. between speech and the linguistic representations that it encodes. 18 BEGINNING CONCEPTS BEGINNING CONCEPTS 19

I J,l1idII 0'0' •.1I!

~I kI > < I I , I , I I I

Figure 1.4 Wavefonn for the sentence Linda loves the melody, illustrating Figure 1.5 Muller-Lyer illusion. In the figure on the left, two horizontal lines graphically the continuous nature of the speech signal. The superimposed ver­ appear to be different lengths, the one on the bottom seeming longer than the tical lines mark the approximate locations for word boundaries. The word one on the top. On the right, the exact copy of that figure demonstrates that the boundaries are not particularly salient, and neither are the boundaries between two horizontal lines are in fact of identical length. the consonants and vowels that make up the words.

A is never consciously available to us; what we are aware of is In fact, even the phonological representation of a sentence is far the mental percept that the stimulus gives rise to. An example of this removed from the properties of the acoustic signal. The phonological process can be illustrated by viewing optical illusions, like that shown representation can be thought of as an idealization of the physical in Figure 1.5. With the Muller-Lyer illusion (Muller-Lyer 1889), in the speech sounds. The abstract representation is made up of discrete left panel of the figure, the stimulus that actually falls on our retinas phonological units (consonants and vowels, syllables, and higher­ contains two horizontal lines of equal length, but we perceive the order rhythmic units, like prosodic words and intonational phrases). bottom line to be longer than the line on the top. The percept of relative The physical signal itself is very different, however. The portions that length depends not just on the actual length of the lines, but also on the correspond to abstract phonological units overlap, and the words run context in which they occur. The fact that these lines are adjoined by together; this is illustrated in Figure 1.4, which shows that the wave­ angles pointing in different directions affects our perceptual interpreta­ form for an utterance is continuous. The speaker may be speaking tion of their length. rapidly and with an unfamiliar accent, with chewing gum in her Perceiving a linguistic representation based on the stimulus of a mouth and with a radio playing in the background, all of which will speech signal requires the hearer to have linguistic competence. affect the signal, making it measurably different from a signal for the Knowledge of language is necessary for a person to reconstruct, and same sentence produced slowly by a native speaker with no gum in therefore perceive, the phonological representation for the speech her mouth and in a quiet room. The relationship between the continu­ signal, which then unlocks the sequence of words and in tum gives ous (and perhaps very noisy) physical signal the hearer receives and way to building the syntactic structure for the sentence. Without lin­ the neatly structured units of the idealized phonological representa­ guistic knowledge, a hearer would be unable to perceive anything other tion he must reconstruct is not at all direct. A complex set of mental than a jumble of disorganized sounds. For example, dogs can be excel­ processing mechanisms must consult the hearer's grammar and lexi­ lent communicators, but they have no knowledge of language, so when con in order to reconstruct a series of linguistic representations, result­ they hear speech, they may recognize the acoustic signal associated ing in the recovery of the speaker's meaning. Researchers think that with their names and a number of familiar commands, but that is all. those mental processes are executed by neurophysiological opera­ Animals "understand" what we say to them through our tone of voice, tions that are specialized for the perception of speech as a linguistic body language, and gaze. For humans, understanding a sentence object. involves very different processes: the organization of sounds, words, In every modality people make the distinction between the actual and ultimately sentences derives from human knowledge of language stimulus (the physical signal) that impinges on our eyes or ears and the and takes on the form of mental representations reconstructed, quite percept that the brain constructs when we interpret that stimulus. indirectly, from the physical speech signal. 1.0 ""I'IINNINO CONfI'PTS BEGINNING CONCEPTS 21

physical speech signal. This view of language was later determined to • Origins of Contemporary Psycho1inguistics be fundamentally flawed, and is diametrically opposed to the view of language presented in this book. Contemporary psycholinguistics is an interdisciplinary field combin­ There were, of course, linguists and who saw difficul­ ing the two disciplines of linguistics and experimental cognitive psy­ ties with the traditional view. The famous linguist Edward Sapir wrote chology. Obviously, this union will be successful only to the extent that a paper entitled "The psychological reality of ," suggesting the two subfields have compatible views of language. When the field of that the mental representation of language should be addressed rather psycholinguistics was first developed, this compatibility was indeed than focus exclusively on its physical representation (Sapir 1949). The the case, just as it is now. What is interesting is that those views have Karl Lashley wrote the now-classic paper "The problem of changed dramatically over the last few decades. serial order in behavior," questioning the explanatory power of associa­ The inception of the field of psycholinguistics occurred in the tive chaining (Lashley 1951). The general view of linguists and psy­ summer of 1951 when, at a meeting of the Research chologists at this time, however, was of language as a system of discrete Council at Cornell University, a committee on Linguistics and behaviors that could be observed, classified, and understood in the Psychology was formed, with Charles Osgood as its chairman (Kess individual as chains of associated behaviors, created by conditioning in 1992). Subsequently, in the summer of 1953, a seminar was held at childhood. These principles of conditioning were taken to be general Indiana University in conjunction with the Linguistic Institute. This principles of learning for all organisms. seminar formed the basis of the first book with psycholinguistics in its This view oflanguage was challenged, beginning in the late 1950s, title, Psycholinguistics: A Survey o/Theory and Research Problems (Osgood by Massachusetts Institute of Technology professor , and Sebeok 1954). At that time linguists focused on a taxonomic analy­ who proposed an entirely new way to think of human language, an sis of languages, which meant that they had as their primary goal the approach that has been adopted by contemporary linguists and psy­ classification of observable aspects of language. When linguists of that chologists, and scholars in related fields, and is essentially the view era approached a new language, their method of analysis was to listen adopted in this book. Chomsky (1959) said that speech should not be to the speakers of the language, figure out what the phonological units the object of study for those who want to understand human lan­ were, and then classify them further into higher-level categories. This guage. Instead, the object of study should be the set of rules - in the method fit in well with the view of language held by psychologists, mind (which is really an abstract term to refer to the brain) - that which was that speech was simply a type of motor behavior exhibited create sentences and underlie observable speech. This is the gram­ by people. The behaviorist psychology of that day took the domain of matical system, and it is not observable in the same way that speech psychology to be behavior (of people or animals), rather than mental is observable (Chomsky 1975). Nonetheless, it is possible to test operations of any kind. They believed that all behaviors could be hypotheses about the properties of the grammatical system and explained as associated (linked) chains of smaller behaviors. Thus, thereby discover the set of rules that constitute people's knowledge speech was regarded as behavioral units of sound combined into of their language. Children acquire language as effortlessly as they words, which were then associated to form phrases, and so on. do, not because there are any general principles ofleaming that apply Acquisition in the child was thought to be the process by which cor­ to all organisms (as argued by behaviorist psychologists), but because rectly associated speech behaviors were built up by rewarding the this internal system of rules is biologically based in the human spe­ desired ones and failing to reward the undesiredsequences. Behaviorists cies (Chomsky 1975). believed that this system of learning, known as conditioning, was Obviously, the Chomskyan conception of language was totally common to all organisms, and that all organisms learned everything incompatible with the behaviorist view. A few psychologists, includ­ the same way. All learning consisted of the acquisition of behavioral ing George Miller (1965), were instantly aware of the implications of routines, and all behavioral routines were acquired by the same princi­ Chomsky'S ideas to the psychological study of language and its acqui­ ples of leaming. The common thread that bound linguistics and psy­ sition. These psychologists were primarily responsible for bringing chology at the middle of the twentieth century, then, was the view that those ideas to the attention of the psychological community. In 1961, everything interesting about language is directly observable in the the linguist Sol Saporta published a volume sponsored by the Social BEGINNING CONCEPTS 23 22 "'UH1,t4INa rnNf'I'PTIi

Science Research Council's Committee on Psychology and Linguistics (just 8 years after the first), titled Psycholinguistics: A Book of Readings; the volume (Saporta 1961) included papers by Chomsky and Miller, as well as by more traditional linguists and psychologists. This publica­ tion ignited an intellectual battle that raged for more than a decade. In the end, an entirely different view of language became accepted, one different from the view that had united linguistics and psychology in the middle of the century. Now, at the onset of the twenty-first century, both fields predominantly accept the Chomskyan view of language as an abstract system represented in the mind or brain that is unique to the human species, develops in the maturing , and underlies but is only indirectly related to physical speech. This is not to say that con­ troversies have disappeared within the fields of linguistics and psy­ chology regarding the best way to characterize linguistic knowledge and investigate its use by and acquisition by children. Healthy fields always contain controversy. However, the controversy exists among people who have the same basic view of language as an object of study.

• How This Book 1s Organized

This book is structured around the issues presented in this chapter. The Study Questions next chapter describes the rules and principles that constitute linguistic 1. What are the two types of linguistic creativity that give us insight competence and the information contained in the lexicon. Chapter 3 presents arguments for the biological basis of language and describes into the nature of human language? its neurological representation. Chapter 4 discusses the process of lan­ 2. Why is it important to distinguish between language and general guage acquisition in the child within the context of a nativist view of intelligence? Between language and communication? ldentiJY language. Chapters 5, 6, and 7 describe the encoding and decoding processes of the speaker and the hearer. Finally, Chapter 8 deals with specific examples. the use of language in memory and . Every chapter has a list of new concepts corresponding to terms in 3. Why are linguists interested in describing rather than prescribing boldface in the body of the text. Every chapter also has a set of study grammar? questions designed to help you read with focus. All references cited in the chapters are listed together at the end of the book. 4. Why might some people think that one speech style or dialect is Throughout the book, we describe a broad range of empirical evi­ better than another? Is this a psycholinguistic issue or a social dence on how language is acquired and processed. We have tried to be issue? Why? explicit in our descriptions of experimental methods when they come up. In recognition that much of this may be new to you, we have also 5. What determines the meaning of a sentence? included an appendix that explains in some detail how psycholinguis­ tic are designed and gives some examples of techniques 6. What does it mean to say that structure is psychologically real, that are commonly in use. though abstract?